Third Genetic Code Anyone?

Abstract

This year is the fiftieth anniversary of the great paper [1Crick F.H.C. Barnett L. Brenner S. Watts-Tobin- R.J. General nature of the genetic code for proteins.Nature. 1961; 192: 1227-1232Crossref PubMed Scopus (551) Google Scholar] describing the general form of the genetic code, a beautiful theoretical and genetic study making a strong case that proteins are encoded in polynucleotides by strings of contiguous, non-overlapping base triplets. This was soon followed by the biochemical work of Marshall Nirenberg and others which elucidated the detailed mapping between triplet codons and amino acids that turns out to be (more or less) universal, a strong testament to the fundamental unity of life. I have noticed that every now and again one sees reference to a “second genetic code”, but why never to a third or fourth genetic code….? Try a Google search using “second genetic code” – there are many hits, but the curious thing is the number of quite distinct second genetic codes that have been claimed. When I try again with “third genetic code” all I get are references to the third base position of the “first” genetic code. Is the fact that the numbering of potential additional genetic codes is stuck on number two perhaps telling us something? In most cases, the claims to have identified a second genetic code have come from work on genuinely interesting biological phenomena. An early claimant was the set of binding specificities of the amino-acyl tRNA transferases, which connect tRNA to their cognate amino acids. These binding specificities are clearly rather special: they are determined of course by the structures of the enzymes, and these structures depend on the enzymes' amino acid sequences and thus on the very genetic code they help to implement. There is a circularity here which is interesting: these enzymes and the tRNAs they charge are the bearers of important “information”. But to refer to this as a second genetic code is confusing, as really it is all part of the molecular basis of the (first) genetic code. Further confusion is generated by another eager candidate: the histone code. The idea here is chemical modifications of histones constitute a kind of code for gene regulation — extreme proponents [2Unsigned editorialTime for the epigenome.Nature. 2010; 463: 587Google Scholar] even assert that this provides “information beyond the genome”, accounting for how complex organisms can be encoded by what in their view are too few genes (particularly in comparison to what the proponents clearly consider significantly simpler organisms). Even ignoring for the moment the lack of any real justification of the latter assumption, the notion that histone modifications represent information beyond the genome is a curious one, when so far as we know the specificity of those modifications depends on genome-encoded proteins. The hope of believers is that, once the modifications have been made they are “self-replicating” — they guide similar modification of daughter chromatin. This remains a possibility, but so far at least there is little hard evidence for such self-replication, independent of sequence-specific DNA-binding proteins or RNAs. Whether or not they are self-perpetuating, histone modifications no doubt play some important role in gene regulation. The regulation of proteins by chemical modification is very widespread of course, and in many cases the modification works by attracting a second protein to the modified one; as, for example, in the case of ubiquitination, which leads to destruction of a protein by the proteasome. In a sense the modification is informational, acting as a tag that says “do this”, rather than working directly by its chemical effects on the modified protein. But I am not sure that this is a code in as clear a sense as the genetic one, where one type of information (a nucleic acid sequence) is converted into another (a protein sequence) — and, importantly, where nucleotide triplets can be strung together to encode a protein of arbitrary length and sequence. The capacity to make arbitrary messages is surely an important feature of a real code. There are other claimants: RNA sequences that determine sites of pre-mRNA splicing; DNA sequences supposed to influence the phasing of nucleosomes along chromatin; the chemical properties of amino acids that influence protein folding; and so on. As each eager new candidate comes along, it is invariably dubbed a new second genetic code — never a third or fourth genetic code…. Why is this? In the contemporary parlance of the internet age, a kind of crowd-sourced opinion is being made, a thumbs down to the claim, which, if truly meaningful and useful, would surely be taken up into general usage, to become the second genetic code. I would suggest we accord the one, universal genetic code its deserved special place by not nominating others to join it in a list.